INDEX

Explanations

phrases related to conflict or violence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the emergence or beginning of conflicts or disturbances

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

antry

-0.79

 confir

-0.73

opa

-0.66

 downgrade

-0.62

 miss

-0.60

 misses

-0.60

oppy

-0.59

 vetoed

-0.59

 retracted

-0.58

osuke

-0.57

POSITIVE LOGITS

stretched

0.82

quished

0.75

olate

0.72

flows

0.71

casts

0.71

rer

0.70

flow

0.68

Sax

0.67

rers

0.66

 valves

0.64

Activations Density 0.025%

phrases related to conflict or violence

phrases indicating the emergence or beginning of conflicts or disturbances

No Comments

No Known Activations